Ontologies COMP6028 Semantic Web Technologies Dr Nicholas Gibbins

Knowledge Representation Knowledge representation is central to the Semantic Web –Data published on the Semantic Web must be structured and organised Long-standing concern in Artificial Intelligence –A good knowledge representation ‘naturally’ represents a given problem domain –A poor knowledge representation is unintelligible

Knowledge Representation Common KR approaches: –Logic –Production rules –Semantic Networks –Frames The Semantic Web combines aspects of all of these schemes

Knowledge Representation Most AI systems (and therefore SW systems) consist of: A knowledge base (KB) –Forms the system's intelligence source –Structured according to the knowledge representation approach taken An inference mechanism –Set of procedures that are used to examine the knowledge base to answer questions, solve problems or make decisions within the domain

Ontologies

Defining the ‘O’ word Ontology, n. 1. a. Philos. The science or study of being; that branch of metaphysics concerned with the nature or essence of being or existence. Oxford English Dictionary, 2004

Defining the ‘O’ word An ontology is a specification of a conceptualisation Specification: A formal description Conceptualisation: The objects, concepts, and other entities that are assumed to exist in some area of interest and the relationships that hold among them Referred to in the philosophical literature as Formal Ontology T. R. Gruber. A translation approach to portable ontologies. Knowledge Acquisition, 5(2): , 1993

Ontology in Computer Science Ontologies as engineered artifacts: –constituted by a specific vocabulary used to describe a certain reality, plus –a set of explicit assumptions regarding the intended meaning of the vocabulary Benefits: –Shared understanding –Facilitate communication –Establish a joint terminology for a community of interest –Normative models –Inter-operability: sharing and reuse

Ontology Structure Ontologies typically have two distinct components: Names for important concepts in the domain –Elephant is a concept whose members are a kind of animal –Herbivore is a concept whose members are exactly those animals who eat only plants or parts of plants –Adult_Elephant is a concept whose members are exactly those elephants whose age is greater than 20 years Background knowledge/constraints on the domain –Adult_Elephants weigh at least 2,000 kg –All Elephants are either African_Elephants or Indian_Elephants –No individual can be both a Herbivore and a Carnivore

Informal Usage Informally, ‘ontology’ may also be used to describe a number of other types of conceptual specification: –Controlled vocabulary –Taxonomy –Thesaurus Study of ontology is not limited to computer scientists and philosophers –Rich tradition of knowledge representation and ontology in library and information science… –…but they talk about classification and metadata instead of ontologies

Controlled Vocabularies An explicitly enumerated list of terms, each with an unambiguous, non-redundant definition No structure exists between terms - a controlled vocabulary is a flat list Examples: –Library of Congress Subject Headings (LCSH) –Medical Subject Headings (MeSH)

Taxonomies A collection of controlled vocabulary terms organised into a hierarchical structure Each term is in one or more parent-child relationships May be several different types of parent-child relationship: –Type-instance –Genus-species –Part-whole (referred to as meronymy)

Taxonomy Examples Library classification schemes –Library of Congress –Dewey Decimal –UDC Linnean Classification –Kingdom, Phylum, Class, Order, Family, Genus, Species, Subspecies MeSH Tree Structures

Taxonomy Examples Dewey Decimal –5xx - Natural Sciences and Mathematics –53x - Physics –537 - Electricity and Electronics Library of Congress –Q - Science –QA - Mathematics –QA Instruments and machines –QA Calculating machines –QA Electronic computers and computer science –QA Computer software

Polyhierachical Taxonomies Define several orthogonal hierarchies –Objects may be classified under multiple hierarchies –Also known as faceted taxonomies Example: Universal Decimal Classification –Facets for language, relation to other subjects – artificial intelligence –616 - clinical medicine –004.8=20 - artificial intelligence in English –004.8:616 - artificial intelligence and clinical medicine –004.8:616=20 - AI and clinical medicine in English

Thesauri A thesaurus is a taxonomy with additional relations showing lateral connections –Related Term (RT) –See Also Parent-child relation usually described in terms of Broader Terms (BT) and Narrower Terms (NT) Thesauri also typically contain scope notes which define the meaning of a term

Thesaurus Example Apples Scope notes:The fruit of any member of the species Malus pumila Broader term: Foodstuffs Related terms: Cooking Ingredients Taxable Foodstuffs Horticulture Narrower terms: Granny Smiths See also: Apple Trees Use: For Apple computers use Personal Computers (Apple)

Ontology An ontology further specialises types of relationships (particularly related term) A ontology typically includes: –Class definitions and hierarchy –Relation definitions and hierarchy An ontology may also include the following: –Constraints –Axioms –Rule-based knowledge

Summary Controlled Vocabulary + Hierarchy = Taxonomy Taxonomy + lateral relations = Thesaurus Thesaurus + typed relations + constraints + rules + axioms = Ontology